Adaptive Significance Levels in Tests for Linear Regression Models: The e-Value and P-Value Cases